1. Field of the Invention
The present invention relates to an imaging apparatus, a control method of an imaging apparatus, and a computer program. More specifically, the invention relates to an imaging apparatus, a control method of an imaging apparatus, and a computer program, which allow an accurate auto-focusing process for a target subject.
2. Description of the Related Art
Nowadays, many imaging apparatus such as a still camera and a video camera have an auto focusing (AF) mechanism mounted thereon which automatically bring a subject into focus. In addition, in recent years, in a camera having an auto focusing (AF) function, such a function, a so-called “Multi AF”, or “multipoint range AF”, is widely adopted, which measures a distance to a plurality of areas in a screen, selects the optimum area resulting from various priorities such as one being closest to a photographer, and one being closer to the center of the screen, and drives a lens to bring the area into focus. This function is applied to allow proper focusing for a main subject with no special operations done by a photographer even though the main subject is not always on the center of the screen.
However, such a Multi AF or multipoint range AF function does not serve for all purposes, and these functions do not always serve for all compositions. In general camera shooting, people are the main subject in a great many cases, but people are at many places in a taken image, and there are various compositions. In the first place, in some cases, a target subject is at the position at which the area to measure a distance is not set in an imaging apparatus. In this case, even though the Multi AF or multipoint range AF function is applied, people are not brought into focus properly.
In order to solve such problems, for example, in Patent Reference 1 (see JP-A-2004-219461), a method is proposed in which skin tone information is detected from an acquired image in an imaging apparatus, the area of the skin tone is assumed as the position of a face for measuring the distance. With this method, such a camera can be implemented that brings a person into focus in any compositions. However, the method has the disadvantage that it is difficult to exert functions well in bright and dark environments where detection performance of the face has little effect.
Furthermore, Patent Reference 2 (see JP-A-2003-107335) discloses a configuration in which a person's face is detected to improve the accuracy of detecting a person, not estimating a face position from skin tone information. In addition to this, it proposes a method in which parts such as a mouth, a nose, and eyes are detected to measure the distance to the parts. However, it is configured in which when a mouth, a nose, and eyes are detected, a user of a camera is allowed to select a part among them, not the camera, which demands user operation and does not implement simple shooting. Moreover, it is likely to lose a chance to take a picture while a user is making a selection of a part.
In the case of the configuration in which a face is detected based on an image acquired by a camera to measure the distance to the face, when the face area in the image acquired by the camera is too small, it causes a problem that this configuration tends to receive influence of an unstable camera and an unstable subject. Moreover, also in the case in which auto-focusing that detects the brightness difference of the face is adapted, there is an essential problem that it is difficult to detect the brightness difference.
For the configuration that solves such problems, Patent Reference 3 (see JP-A-2004-317699) proposes a method that changes a measuring area in accordance with the size of a face. It is the method that in the case in which a face is larger than a predetermined size and eyes are detected, the area that includes the eyes therein is set to a measuring area, and in the case in which a face is smaller than a predetermined size, a large frame that surrounds the face is set. However, according to the method, in the case in which zooming is used for shooting when a face is small, such a problem arises that the background comes into focus and the face is out of focus.